Phosphinic Acid-Based MMP-13 Inhibitors That SpareMMP-1 and MMP-3

Lawrence A. Reiter,* Peter G. Mitchell, Gary J. Martinelli, Lori L. Lopresti-Morrow,Sue A. Yocum and James D. Eskra

Pfizer Inc, Global Reseach & Development, Groton Laboratories, Eastern Point Rd., Groton, CT 06340, USA

Received 19 February 2003; revised 17 April 2003; accepted 22 April 2003

Abstract—Phosphinic acid-based inhibitors of MMP-13 have been investigated with the aim of identifying potent inhibitors withhigh selectivity versus MMP-1. Independent variation of the substituents on a P1

0 phenethyl group and a P2 benzyl group improvedpotencies in both cases around 3-fold over the unsubstituted parent. Combining improved P1

0 and P2 groups into a single moleculegave an inhibitor with a 4.5 nM IC50 against MMP-13 and which is 270-fold selective over MMP-1.# 2003 Elsevier Science Ltd. All rights reserved.

The quest to identify therapeutically useful inhibitors ofMMP’s has entered its third decade1, and this questremains to be fulfilled. The early studies, which wereinitiated in the pre-genomics era, aimed at inhibitingMMP-1 and focused on identifying molecules whichwere potent and which had favorable pharmacokinetic(PK) properties. These early endeavors have beenreplaced with more fundamental work aimed at identi-fying the many MMP’s present in humans, assessingtheir relative abundance and distribution, and most cri-tically, determining their relevance to different diseasestates.2 The search for therapeutically useful moleculeshas likewise evolved. Early studies focused on peptidicstructures some of which displayed potent MMP-1inhibition but which had unknown (and un-appreciatedlack of) specificity and poor PK properties. While pro-gress has been made in identifying compounds withimproved PK in both peptidic and non-peptidic motifs,clinical experience with broad spectrum inhibitors hasrevealed a common, intolerable side effect, musculo-skeletal syndrome (MSS).3 Avoiding MSS is likely torequire identification of the MMP(s) responsible for theside effect and the discovery and development ofappropriately selective inhibitors which in an ideal case,would inhibit only the target MMP(s).

That MMP-1 plays a role in the development of MSShas been postulated4 and efforts to identify MMPinhibitors that spare MMP-1 have been reported.4,5 Thetarget MMPs for these studies, though, have varieddepending on which disease is being studied. In the caseof osteoarthritis, the presence of MMP-13 in OA carti-lage tissue in humans,6,7 its ability to turn over type IIcollagen,6,8 its co-localization with cleaved type II col-lagen in OA cartilage tissue,9 and it degradative prop-erties in cartilage,10 all point to its likely role in thepathology of the disease.11 Additional evidence thatMMP-13 is a key driver of cartilage destruction (vsMMP-1) was provided by the study of a relatively broadspectrum inhibitor, CGS 27023A,12 versus 1, a MMP-1-sparing MMP-13 inhibitor, in the bovine nasal cartilage(BNC) model of cartilage degradation (Fig. 1).13 In thismodel CGS 27023A displayed an IC50 for inhibition ofhydroxyproline release of �50 nM which is about 10-fold higher than its IC50 against recombinant MMP-13(3 nM) and 3-fold higher than its MMP-1 IC50 (18 nM).The phosphinate 1 displayed an IC50 in the BNC modelof �500 nM, about 10-fold higher than its MMP-13IC50 (70 nM), consistent with the result with CGS27023A. Furthermore, the IC50 of 1 in the BNC modelis about 40-fold less than its MMP-1 IC50 (22,000 nM)clearly demonstrating that MMP-1 does not play sig-nificant role in this model.

We recently reported a series of phosphinate-basedinhibitors, including 2, the t-butylglycine-containinganalogue of 1 (MMP-13 IC50 30 nM, MMP-1 IC50

0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.doi:10.1016/S0960-894X(03)00413-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2331–2336

*Corresponding author. Tel.:+1-8604-413-681; fax:+1-8606-861-176;e-mail: lawrence_a_reiter@groton.pfizer.com

>30,000 nM), that take advantage of binding intopockets on both sides of the scissile bond.14 In parti-cular, key interactions occur in the S2 and S1

0 pocketsand with the catalytic zinc atom. The premise of thiswork was that, through the interaction with more sub-sites of the enzyme, the identification of selective com-pounds may be possible. Since our group is focused onpotential treatments for osteoarthritis, we wished tocontinue to explore this premise in the context of MMP-13 inhibition. We especially sought compounds with>100-fold selectivity versus MMP-1. The most potentMMP-13 inhibitor identified in our previous work was3, having an MMP-13 IC50 of 21 nM and 37-fold selec-tivity over MMP-1 (Table 1). Although 2 was moreselective versus MMP-1, the somewhat greater potencyof 3 and the synthetic accessibility and lower molecularweight of P1

0 phenethyl derivatives led us to explore thelatter sub-series. Herein we report the results of our stud-ies in which we have varied the P2 and P1

0 substituentsseeking potent and selective MMP-13 inhibitors.

The synthesis of P10 analogues was accomplished in 4

synthetic steps (Scheme 1). Appropriately substitutedethyl a-phenethylacrylates were prepared through reac-tion of various benzyl cuprates with ethyl a-bromo-

methylacrylate.15 The use of an excess of cuprate drovethe reaction to completion leading to a crude productcontaining the desired acrylate and the substituted tol-uene derived from the quenched cuprate. This latter,neutral impurity was easily removed after the next syn-thetic step in which the mono-substituted phosphinicacid 4, prepared through an Arbuzov reaction of thecorresponding benzyl bromide with bis(trimethyl-silyl)phosphonite,16 was reacted with each a-substitutedacrylate.17 The crude di-substituted phosphinic acid waspurified by absorption onto MP-carbonate (ArgonautTechnologies) resin, washing to remove non-acidiccomponents and release of product with TFA inCH2Cl2. The phosphinates thus prepared were pure byNMR and LC-MS. Subsequent saponification andselective coupling of the carboxylic acid to S tert-leucineN-methylamide using PyBOP18 gave the crude desiredcompounds as a mixture of diastereomers. PreparativeRP (reverse phase) chromatography yielded the purecompounds 5–13.

The synthesis of the P2 analogues made use of inter-mediate 14 in order to facilitate the rapid preparation ofthe target compounds (Scheme 2). This was preparedfrom benzyl a-phenethylacrylate and 4-iodobenzyl-bromide as previously described.14 The intermediatemono-substituted phosphinic acid adduct of bis(tri-methylsilyl)phosphonite and benzyl a-phenethylacrylatewas partially resolved (�3/1 S/R) by one crystallizationof its S a-methylbenzylamine salt.19 Negishi coupling of14 with benzyl zinc reagents20 gave the intermediatescontaining the various P2 substituents. Conversion intothe target phosphinic acids 15–27 was achieved as pre-viously described.14 Owing to the partial resolution ofthe earlier intermediate, the diastereomeric mixtureswere all biased toward the more potent isomer. Never-theless, if the diastereomeric mixture was sufficientlypotent, the pure isomer was isolated by preparative RPchromatography (e.g., 15, 18, 21, and 26).

The analogue combining the better P10 and P2 sub-

stituents, 28, was prepared from 4-(2-methoxy)benzyl-benzyl bromide and ethyl a-(4-chlorophenyl)ethylacrylateas described above for the other P1

0 analogues.

Substitution on the terminus of the phenethyl P10 group

modulated MMP-13 potency and/or selectivity versusMMP-1. In all cases but one, substitution either had aslight effect (e.g., 6) or modestly increased MMP-13potency from 2- to 4-fold. The exception, the 2-methylderivative 8, was substantially less active. Althoughincreases in potency were only modest at best, these,coupled with modest reductions in MMP-1 potency, ledto selectivities of greater than 100-fold for some of the4-substituted derivatives. The loss of MMP-1 potencyfor compounds with additional substitution at C-4 ofthe phenylethyl group is consistent with the MMP-1 S1

0

pocket being shallow.

In contrast to the considerable published SAR for P10

substitution of MMP-13 inhibitors,2 little is knownabout the S2 pocket of MMP’s.

21 As a result, we choseto initially survey the effect of fluoro, methyl and

Figure 1. Inhibition of IL-1 induced degradation of bovine nasal car-tilage (BNC) explants.

2332 L. A. Reiter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2331–2336

methoxy groups at the 2-, 3- and 4-positions of theterminal P2 benzyl group. In contrast to the findingswith P1

0 substituents, most such substitutions led todecreases in activity. However, within each substituentset, the 2-substituted derivatives were the most potentwith the 2-methoxy analogue 21 being both more potentand more selective than 3. The observed increase inselectivity of 21 derives from its increased MMP-13potency (3-fold) as MMP-1 potency remained the same.This suggested that a 2-substituent may be interacting ina favorable way with the S2 pocket of MMP-13 whileenjoying little additional positive interaction with the S2pocket of MMP-1. In order to possibly capitalize onthis, we prepared additional 2-substituted derivatives,24–27. However, none of these were more potent orselective than 21.

Having defined improved substituents for both the P2benzyl group and the P1

0 phenethyl group, we then pre-pared 28 which combines these substituents—a 2-meth-oxybenzyl group at P2 and a 4-chlorophenethyl group atP1

0—in a single molecule. As we had hoped, the sub-stituents in this compound did display modest syner-gism; excellent potency (MMP-13 IC50 4.5 nM) andhigh selectivity versus MMP-1 (270-fold) was observed.

Having achieved our key goal of identifying compoundswith >100-fold selectivity over MMP-1, we next exam-ined the selectivity of this series of compounds versusother MMP’s. All of the compounds in Table 1 weretested against MMP-2 and MMP-3; however, only datafor selected compounds, 3, 11, 21 and 28, are presentedin Table 2; these data are representative of what wasfound for the entire series. The compounds were ingeneral very selective for MMP-13 versus MMP-3 withvirtually all analogues being 100-fold or more selective.On the other hand, selectivity versus MMP-2 wasabsent. In fact, most of these compounds were morepotent against MMP-2 than against MMP-13. In ourearlier report,14 we argued that the difference in poten-cies of 3 and 29, a P2 ‘debenzyl’ analogue, indicated thatthe P2 benzyl group is interacting with the S2 pocket ofMMP-13. Similarly, the MMP-2 potencies of 3 and 29vary 11-fold suggesting that, as with MMP-13, the P2benzyl group is interacting with the S2 pocket (oranother adjacent region) of MMP-2. To broaden theprofile of the series, selected compounds were alsoscreened against MMP-8 and MMP-12, two other deeppocket MMPs. As with MMP-2, selectivity for MMP-13inhibition over either MMP-8 or MMP-12 was notobserved (Table 2). Also, like MMP-2, the �10-fold

Table 1. Variation of P10 and P2 residues. MMP-13 and MMP-1 activity

CPDa Y X MMP-1b MMP-13c MMP-1/ MMP-13 Ratio of S/R isomers

IC50 (nM)

d

IC50 (nM)d IC50 ratio at P1

0 side chain

3

H– H– 770�300 21�9 37 100/0 5 H– 2-F– 280�40 10�3 28 99/1 6 H– 3-F– 860�190 19�6 45 100/0 7 H– 4-F– 530�110 13�3 41 100/0 8 H– 2-Me– 12,000�900 >300 (3) <40 100/0 9 H– 3-Me– 120�10 4.3�0.8 28 100/0 10 H– 4-Me– 630�30 7.7�2.4 82 100/0 11 H– 4-Cl– 1100�200 8.3�2.8 130 100/0 12 H– 4-Br– 1400�100 9.4�1.2 150 100/0 13 H– 4-MeO– 1600�200 7.7�2.0 210 100/0 15 2-F– H– 690�320 18�7.8 38 96/04 16 3-F– H– 1600�300 58�41 28 75/25 17 4-F– H– 970�50 44�31 22 74/26 18 2-Me– H– 1100�500 24�21 46 96/04 19 3-Me– H– 2300�1000 37�13 62 80/20 20 4-Me– H– 1600�300 29�9 55 79/21 21 2-MeO– H– 770�290 7.0�3.8 110 99/01 22 3-MeO– H– 3000�800 43�18 70 78/22 23 4-MeO– H– 5200�500 54�23 96 79/21 24 2-Cl– H– 430�190 17�7 25 86/14 25 2-Et– H– 350�40 13�7 27 86/14 26 2-EtO– H– 540�190 11�4 49 98/02 27 2-CF3– H– 650�370 31�23 21 73/23 28 2-MeO– 4-Cl– 1200�10 (2) 4.5�2.2 270 100/0

aAll compounds were characterized by 1H NMR, MS and HPLC. The latter assured purity and allowed the diastereomeric ratio, if any, of the P10

side chain to be determined.bDetermined using full-length MMP-1 by the method of Bickett (ref 22).cDetermined using full-length MMP-13 by the method of Bickett (ref 22).dValues are the mean�S.D. of 3 determination unless otherwise noted (n).

L. A. Reiter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2331–2336 2333

shift in potency between the ‘de-benzyl‘ analogue 29 and3 suggests that the inhibitors are interacting in the S2pocket as well as the S1

0 pocket.

Through systematic variation of substituents we haveimproved the potency of phosphinate-based MMP-13inhibitors over those reported earlier. Furthermore,high selectivity between MMP-13 and MMP-1 has beendemonstrated. With the exception of MMP-3, selectivity

against other MMPs was not observed. However,comparisons of MMP-2, -8, -12 and -13 potencies forthe ‘de-benzyl’ analogue 29 versus 3 (�10-fold differ-ence in all cases) suggests that key interactions areoccurring with both the P2 and P1

0 groups. Therefore,further manipulation of these groups may lead to com-pounds selective towards these other enzymes. Selectiv-ity against these particular MMP’s may or may notultimately be sought depending on whether selectivity

Scheme 2.

Scheme 1.

2334 L. A. Reiter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2331–2336

versus MMP-1 is sufficient to avoid MSS or on whatother MMPs are eventually identified as being keycontributors to this undesired side effect.

References and Notes

1. Three decades have elapsed from the first publications ofinhibitors that took advantage of specific interactions toMMP-1 sub-sites: Gray, R. D.; Saneii, H. H.; Spatola, A. F.Biochem. Biophys. Res. Commun. 1981, 101, 1251. Sundeen, J.E.; Dejneka, T. U.S. Patent 4 235 885, Nov. 25, 1980.2. See the following reviews for changing trends over the past�30 years: Johnson, W. H.; Roberts, N. A.; Borkakoti, N. J.Enzyme Inhib. 1987, 2, 1. Henderson, B.; Docherty, A. J. P.;Beeley, N. R. A. Drugs Future 1990, 15, 495. Zask, A.; Levin,J. I.; Killar, L. M.; Skotnicki, J. S. Curr. Pharm. Des. 1996, 2,624. Whittaker, M.; Floyd, C. D.; Brown, P.; Gearing, A. J. H.Chem. Rev. 1999, 99, 2735. Skiles, J. W.; Gonnella, N. C.;Jeng, A. Y. Curr. Med. Chem. 2001, 8, 425.3. Hutchinson, J. W.; Tierney, G. M.; Parsons, S. L.; Davis,T. R. C. J. Bone Joint Surg. (Br) 1998, 80-B, 907. Wojtowicz-Praga, S.; Torri, J.; Johnson, M.; Steen, V.; Marshall, J.; Ness,E.; Dickson, R.; Sale, M.; Rasmussen, H. S.; Chiodo, T. A.;Hawkins, M. J. J. Clin. Oncol. 1998, 16, 2150.4. Freskos, J. N.; McDonald, J. J.; Mischke, B. V.; Mullins,P. B.; Shieh, H.-S.; Stegeman, R. A.; Stevens, A. M. Bioorg.Med. Chem. Lett. 1999, 9, 1757.5. Freskos, J. N.; Mischke, B. V.; De Crescenzo, G. A.;Heintz, R.; Getman, D. P.; Howard, S. C.; Kishore, N. N.;McDonald, J. J.; Munie, G. E.; Rangwala, S.; Swearingen,C. A.; Voliva, C.; Welsch, D. J. Biorg. Med. Chem. Lett. 1999,9, 943. Barta, T. E.; Becker, D. P.; Bedell, L. J.; De Crescenzo,G. A.; McDonald, J. J.; Munie, G. E.; Rao, S.; Shieh, H.-S.;Stegeman, R.; Stevens, A. M.; Villamil, C. I. Bioorg. Med.Chem. Lett. 2000, 10, 2815. Barta, T. E.; Becker, D. P.; Bedell,L. J.; De Crescenzo, G. A.; McDonald, J. J.; Mehta, P.;

Munie, G. E.; Villamil, C. I. Bioorg. Med. Chem. Lett. 2001,11, 2481. Becker, D. P.; Barta, T. E.; Bedell, L. J.; De Cres-cenzo, G. A.; Freskos, J.; Getman, D. P.; Hockerman, S. L.;Li, M.; Mehta, P.; Mischke, B.; Munie, G. E.; Swearingen, C.;Villamil, C. I. Bioorg. Med. Chem. Lett. 2001, 11, 2719.Becker, D. P.; De Crescenzo, G. A.; Freskos, J.; Getman,D. P.; Hockerman, S. L.; Li, M.; Mehta, P.; Munie, G. E.;Swearingen, C. Bioorg. Med. Chem. Lett. 2001, 11, 2723.6. Mitchell, P. G.; Magna, H. A.; Reeves, L. M.; Lopresti-Morrow, L. L.; Yocum, S. A.; Rosner, P. J.; Geoghegan,K. F.; Hambor, J. E. J. Clin. Invest. 1996, 97, 761.7. Yocum, S. A.; Lopresti-Morrow, L. L.; Reeves, L. M.;Mitchell, P. G. Ann. N.Y. Acad. Sci. 1999, 878, 583. Borden,P.; Solymar, D.; Sucharczuk, A.; Lindman, B.; Cannon, P.;Heller, R. A. J. Biol. Chem. 1996, 271, 23577.8. Knauper, V.; Lopez-Otin, C.; Smith, B.; Knight, G.; Mur-phy, G. J. Biol. Chem. 1996, 271, 1544.9. Freemont, A. J.; Byers, R. J.; Taiwo, Y. O.; Hoyland, J. A.Ann. Rheum. Dis. 1999, 58, 357.10. Billinghurst, R. C.; Dahlberg, L.; Ionescu, M.; Reiner, A.;Bourne, R.; Rorabeck, C.; Mitchell, P.; Hambor, J.; Diek-mann, O.; Tschesche, H.; Chen, J.; Van Wart, H.; Poole, A. R.J. Clin. Invest. 1997, 99, 1534. Billinghurst, R. C.; Wu, W.;Ionescu, M.; Reiner, A.; Dahlberg, L.; Chen, J.; Van Wart, H.;Poole, A. R. Arthritis Rheum. 2000, 43, 664.11. See also ref. in Henrotin, Y.; Sanchez, C.; Reginster, J.-Y.Exp. Opin. Ther. Patents 2002, 12, 29.12. MacPherson, L. J.; Bayburt, E. K.; Capparelli, M. P.;Carroll, B. J.; Goldstein, R.; Justice, M. R.; Zhu, L.; Hu, S. i.;Melton, R. A.; Fryer, L.; Goldberg, R. L.; Doughty, J. R.;Spirito, S.; Blancuzzi, V.; Wilson, D.; O’Byrne, E. M.; Ganu,V.; Parker, D. T. J. Med. Chem. 1997, 40, 2525.13. Mitchell, P. G.; Lopresti-Morrow, L. L.; Yocum, S. A.;Sweeney, F. J.; Reiter, L. A. Ann. N.Y. Acad. Sci. 1994, 732,395.14. Reiter, L. A.; Rizzi, J. P.; Pandit, J.; Lasut, M. J.; McGa-hee, S. M.; Parikh, V. D.; Blake, J. F.; Danley, D. E.; Laird,E. R.; Lopez-Anaya, A.; Lopresti-Morrow, L. L.; Mansour,M. N.; Martinelli, G. J.; Mitchell, P. G.; Owens, B. S.; Pauly,T. A.; Reeves, L. M.; Schulte, G. K.; Yocum, S. A. Bioorg.Med. Chem. Lett. 1999, 9, 127.

Table 2. MMP-13, MMP-3, MMP-2, MMP-8 and MMP-12 activity of selected phosphinic acids

MMP-13a MMP-3b MMP-2c MMP-8d MMP-12e

(IC50 nM)f

(IC50 nM)

f

(IC50 nM)f (Ki nM)

f

(IC50 nM)f

3

21�9 1600�500 10�4 0.77�0.44 8.2�2.5 11 8.3�2.8 900�100 2.6�0.6 0.56�0.10 1.7�0.7 21 7.0�3.8 800�200 6.6�2.4 2.1�0.1 4.0�0.8 28 4.5�2.2 1600�400 6.6�0.5 2.4�0.4 5.0�4.2 29g 210�80 2700 (1) 110 (1) 12�2 93�12 Ratio of IC50’s 29 versus 3

10

— 11 16 11

aDetermined using full-length MMP-13 by the method of Bickett (ref 22).bDetermined using full-length MMP-3 by the method of Nagase (ref 23).cDetermined using full-length MMP-2 by the method of Knight (ref 24).dDetermined using the catalytic domain of MMP-8 by the method of Bickett (ref 22).eDetermined using the catalytic domain of MMP-12 by the method of Bickett (ref 22).fValues are the mean�SD of 3 determinations unless otherwise noted (n).gCompound 29, reported in our earlier work (ref 14), is a 62/38 mixture of S/R P1

0 isomers.

L. A. Reiter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2331–2336 2335

15. Berk, S. C.; Knochel, P.; Yeh, M. C. P. J. Org. Chem. 1988,53, 5789. Jubert, C.; Knochel, P. J. Org. Chem. 1992, 57, 5425.16. Boyd, E. A.; Regan, A. C.; James, K. Tetrahedron Lett.1994, 35, 4223.17. Boyd, E. A.; Corless, M.; James, K.; Regan, A. C. Tetra-hedron Lett. 1990, 31, 2933.18. Campagne, J.-M.; Coste, J.; Guillou, L.; Heitz, A.; Jouin,P. Tetrahedron Lett. 1993, 34, 4181.19. Caldwell, C. G.; Sahoo, S. P.; Polo, S. A.; Eversole, R. R.;Lanza, T. J.; Mills, S. G.; Niedzwiecki, L. M.; Izquierdo-Martin, M.; Chang, B. C.; Harrison, R. K.; Kuo, D. W.; Lin,T.-Y.; Stein, R. L.; Durette, P. L.; Hagmann, W. K. Bioorg.Med. Chem. Lett. 1998, 6, 323.20. Negishi, E. I.; King, A. O.; Okukado, N. J. Org. Chem.1977, 42, 1821.21. Vassiliou, S.; Mucha, A.; Cuniasse, P.; Georgiadis, D.;Lucet-Levannier, K.; Beau, F.; Kannan, R.; Murphy, G.;

Knauper, V.; Rio, M.-C.; Basset, P.; Yiotakis, A.; Dive, V. J.Med. Chem. 1999, 42, 2610. Deng, S.-J.; Bickett, D. M.;Mitchell, J. L.; Lambert, M. H.; Blackburn, R. K.; Carter,H. L., III; Neugebauer, J.; Pahel, G.; Weiner, M. P.; Moss,M. L. J. Biol. Chem. 2000, 275, 31422.22. IC50’s or Ki’s determined using Dnp-Pro-Cha-Gly-Cys(Me)-His-Ala-Lys-(Nma)-NH2 as described by: Bickett,D. M.; Green, M. D.; Berman, J.; Dezube, M.; Howe, A. S.;Brown, P. J.; Rothand, J. T.; McGeehan, G. M. Anal. Bio-chem. 1993, 212, 58.23. IC50’s determined using Mca-Arg-Pro-Lys-Pro-Val-Glu-Nva-Trp-Arg-Lys-Dnp-NH2 as described by: Nagase, H.;Fields, C. G.; Fields, G. B. J. Biol. Chem. 1994, 269,20952.24. IC50’s determined using Mca-Pro-Leu-Gly-Leu-Dpa-Ala-Arg-NH2 as described by:Knight, C. G.; Willenbrock, F.;Murphy, G. FEBS Lett. 1992, 296, 263.

2336 L. A. Reiter et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2331–2336